Qwen-VL-Chat visual part

[Leminhbinh0209]

起始日期 | Start Date

No response

实现PR | Implementation PR

No response

相关Issues | Reference Issues

No response

摘要 | Summary

Hi,

Did you make any modification to visual part in the file modeling_qwen.py. As I understand that you modified the encode method in visual.py file to receive image instead of its path. Could you help me locate where the visual embedding is forward to the LLM model during training?

Thanks

基本示例 | Basic Example

(No)

缺陷 | Drawbacks

(No)

未解决问题 | Unresolved questions

No response